In this study our aim was to answer several questions on which most pediatri cians are interested in. They were as follows:
1. How common are the infections of the central nervous system in Estonia and which are the predominant etiological agents?
2. Is there any additional possibility except common tests to distinguish between bacterial and aseptic meningitis?
3. Would it be possible to prognosticate the outcome already on admission?
4. How long should be a child with BM treated with the antibiotics?
5. Do we need to follow up all children after BM or only some and if we do, which of them should be followed up?
Our retrospective epidemiological study based on annual reports of the hos
pitals where children with ICNS were treated during the study period (from 1980 to 1989) revealed an incidence rate of ICNS 66.2/100 000 children which was very similar to that reported from Spain (Roca et al. 1992). We found it in
teresting that the incidence rates (29.8 and 25.8/100 000 children respectively) as well as the epidemic course of AM and BM were rather similar. We found the etiological pattern of BM completely different from that in Nordic countries, USA and Australia where the leading cause of community acquired meningitis in children is H. influenzae (Wenger et al. 1990; Peltola et al. 1990;
Hanna et al. 1991). Our data with the predominance of N. meningitidis are very similar to studies performed in the former Soviet Union (Djomina ei al. 1985;
Cibiras et al. 1986; Kostjukova et al. 1992). Due to the fact that in more than half of the cases the etiology of BM remained unknown, we can speculate, that under diagnosing of Hib meningitis was going on. With the improvment of facilities of bacteriological laboratories and introducing new methods the number of unidentified cas,es decreased, but meningococci were still most frequently found. So we believe that N. meningitidis is still predominant agent in the etiology of BM in children but like Iljina et al. (1990) we also saw that the importance of Hib raised to the end of the studied period.
Unfortunately the etiology of AM remained unknown in more than 90% of cases, whereas with good laboratory facilities it could be proved in more than 3/4 of cases (Tardieu et al. 1986; Koskiniemi et al. 1989). The main reason for high number of unidentified cases seems to be that AM as a mild disease was mostly treated at district hospitals with the poor diagnostics facilities and low interest of practising doctors for AM etiology. We feel that the doctors interest will increase if rapid methods will become available.
Like other studies (Nicolosi et al. 1986; Etter et al. 1991) we showed the relatively good outcome of AM, whereas the prognosis of BM even at the end of the study period when most of children were referred to Tartu University Children's Hospital was serious. The mortality rate of BM was 13.0% and dis
ability rate 8.9%. The worst outcome with the disability and mortality of 53.8%
and 45.8% respectively was noted like by others after neonatal meningitis (Bell
et al. 1989; Zaki et al. 1990). The reasons of extremely bad outcome of neonatal meningitis are not clear thus far. We only can speculate that it is connected with the immature immune system, but also usually with late diagnosis while there are no characteristic symptoms for meningitis in that age (Schaad et al.
1992). We also saw that during the study period the number of neonatal meningitis cases increased and it could be connected to the opening of NICU, concentrating patients there and spreading of nosocomial infection. Therefore we agree with Schaad et al. (1992) that the most important factor in preventing poor outcome will be profylaxis of neonatal infection.
Despite the fact that from 22 studies only 13 showed that prognosis of BM patients depends on the time of onset of antibacterial therapy (Radetsky 1992) our study demonstrated that duration of illness before hospitalisation was the single clinical characteristic being prognostically important. However, in some cases it will be extremely difficult to distinguish between BM and AM using common tests of CSF (Rodewald et al. 1991). Our work based on the different pathogenesis of BM and AM, measured the activity/concentration of some in tracellular enzymes (AST, GGT, LDH, CPK and CK-BB) to differentiate be
tween BM and AM and tried to found out their prognostic value. We like Sirkis (1982); Agrawal et al. (1989) and Nand et al. (1992) found that patients with BM had significantly higher activity of AST and GGT than those with AM and meningism, however the activity of AST didn't increase the reference levels.
We also showed as Agrawal et al. (1989) and Roslõi et al. (1991) the prognostic value of raised AST activity. Due to the fact that we could not find any correlation between the activity of AST in CSF and blood serum we suggest that local releasing of AST is going on and its raised activity most probably shows the cytolysis of the brain tissue. The benefit of measuring LHD activity in differentiation of AM and BM was also very poor in this study (Landaas et al. 1985), but a good correlation between bad prognosis and high LDH activity on admission was found. It was a surprise that the benefit of measuring of CK-BB was poor, while the CK-BB seems to be one of the best markers in the estimation of brain damage that we have (Beiton et al. 1970;
Sööt et al. 1989; Praeter et al. 1991). In contrast to Haldre et al. (1991) and Talvik (1992) who found good prognostic value of CK-BB in the CSF in patients with head injury and hypoxic-ischaemic brain damage respectively, we found even lower concentration of CK-BB in the CSF in children who either died or recovered with permanent sequelae on admission. The same was also shown by Briem (1982) in patients with meningoencephalitis and we speculate that the prognosis of BM patients depends not only on the extent of brain injury but also on the site of damage. The ground of pathologic process is diffuse vasculitis of the brain (Quagiliarello et al. 1992), but some areas are more affected than others. It was demonstrated by Delanghe et al. (1990) that the distribution of CK-BB in the brain is different, the highest concentrations were found in cortex and Capsula interna, while only limited amounts were measured in pons, cerebellum and medulla oblongata. So we suggest that the importance of estimation of intracellular enzymes in differential diagnosis of meningitis
18 69
can not be overestimated, nevertheless they can give us some additional information in understanding the pathogenesis of BM and AM.
The data of the duration of antibacterial therapy vary largely in the literature and there are few controlled clinical trials thus far. According to the fact that the CSF is sterile in meningococcal meningitis mostly after 24 hours of antibacterial therapy and by Hib and pneumococcal meningitis after 72 hours we decided to use five-day therapy (Peltola et al. 1989; Lebel et al. 1989).
From 42 patients who met the inclusion criteria 36 completed the study according to the study protocol. Main reasons for extended therapy were either prolonged fever, persistent pleocytosis or complicated course of meningitis.
Although neither the persistent pleocytosis nor fever are indications for continuing AB therapy (Chartrand et al. 1976; La Via et al. 1992) it is not possible to complete the AB therapy without special investigations. There was no differences in outcome at the time of hospital discharge and one year after acute illness between children treated for five days and longer. However, the patients who got the five-day therapy were hospitalised for significantly shorter periods and had no hospital acquired respiratory infections. The disadvantage of this study is that it was not a randomised study and historical comparison group was used as a control. It would be not possible to set up a randomised, double-blind study in any country while this would require e.g., approximately 1,400 patients to be followed for 5 years in order to detect a 50% increase in a sequelae with an 80% of certainty (Helwig 1992). Therefore the experience of several small studies could be valuable in elaborating the general rules. Our data showed that five-days of AB therapy would be sufficient for noncomplica ted meningococcal meningitis, but due to the small number of patients we cannot recommend it for patients with pneumococcal and Hib meningitis.
However, we still believe that in patients treated for five days the second lumbar tap is necessary.
The fact that no antibacterial therapy is needed after fifth day does not mean the child's complete recovery. We like Klein et al. (1986) and Lebel et al.
(1989) showed that at the time of hospital discharge 28.5% of patients had changes of muscle tones and 14.3% had ataxia. Both defects were reversible and during the long term follow-up they persisted in only 11.9% and 3.5% of children respectively and were in most cases of mild degree. Our enzyme studies showed also that at the end of antibacterial therapy the activity of CPK and the concentration of CK-BB was still over the reference value which most probably speaks of cytolysis going on. There are no studies published thus far concerning the treatment of BM meningitis if the CSF is sterile and we doubt wheather any medicine will be needed at all. Nevertheless, for patients with permanent defects (hearing disturbances and severe motor abnormalities) rehabilitation had to be started as early as possible. Our study found a highly significant correlation between severe neurological defects at the time of hospital discharge and motor abnormalities and hearing disturbances during the long term follow-up investigation, so we concluded that following up was needed only for those children. We like Annegers at al. (1988) demonstrated that the BM patients have a higher risk later epilepsy. Nevertheless, it was not
possible to prognosticate the development of late seizures, while there was no difference neither in clinical symptoms nor in laboratory tests during the acute illness and at the time of hospital discharge.
As in most studies (Baraff et al. 1993) the most common sequelae after BM were hearing disturbances, found in 8.3% of children during the long-term fol-low-up investigations. Although we didn't find any differences between lan guage disorders and hearing defects however^ we as Brookhauser et al. (1989) believe that hearing disorders should be diagnosed immediately after the acute illness and early rehabilitation should be started. The fact that in two children ataxia was combined with the hearing disturbances supports the opinion of Harada et al. (1988); Kaplan et al. (1989) and Bhatt et al. (1992) that they are results of labyrinthitis. However, the significant correlation between hearing disturbances and motor defects supports the data of Jiang et al. (1990) that is-chaemia in the area of auditory tract may be also important for the development of hearing disturbances.
In conclusion this study of more than ten years demonstrated that although the annual number of patients with ICNS in South-Estonia is not high perma
nent sequelae and deaths still occur. Therefore patients with BM, complicated forms of AM and encephalitis should be concentrated to the hospitals with better laboratory facilities, possibilities for sophisticated methods of intensive care and with the doctors experienced in the care of such patients.
CONCLUSIONS
1. Our data revealed a constant incidence rate of infections of the central ner
vous system of 66.2/ 100 000 children in the 1980s in South-Estonia, which in
cludes 25.8 for bacterial meningitis, 29.8 for aseptic meningitis, 7.6 for tick-borne encephalitis and 3.0/100 000 children for encephalitis caused by other viruses.
2. Bacterial meningitis in contrast to the aseptic meningitis was predominantly a disease of the early childhood — 84.3% of the patients with bacterial meningitis were younger than five years whereas among aseptic meningitis and encephalitis patients more than 80% were over 5 years.
3. Neisseria meningitis which caused 34.9% of bacterial meningitis cases during the study period was the leading micro-organism among the etiologic factors. The importance of Haemophilus influenzae in the etiology of bacterial meningitis increased from 0.9% at the beginning of the 1980s to 8.9% at the end of it.
4. A significant rise in AST and GGT activities in CSF was found in patients with bacterial meningitis during the first days of illness. The significantly in
71
creased activities of AST and LDH could be an indicator of the extent of brain damage and therefore prognostically important.
5. For children with non complicated cases of meningococcal meningitis a five-day antibacterial therapy will be adequate, but due to a small number of patients we can not recommend it for Hib and pneumococcal meningitis.
6. Motor abnormalities, hearing impairment, late seizures and cognitive disor ders have been the most common long-term sequelae of bacterial meningitis found out in 20% of children.
7. Motor defects detected soon after acute illness in 27.3% of patients had ten
dency to resolve with the time and were seen during the long term follow-up examination in only 13.1% of children.
7. There was a significant correlation between different sequelae (but for epilepsy) during the long term follow-up and severe neurologic defects at the time of hospital discharge. It means that only children with severe neurological deficits are at a higher risk for permanent neurological sequelae and hearing disturbances. Those showing no or only mild disturbances have an excellent chance of escaping serious sequelae except epilepsy.
8. The' mortality rate of ICNS was 5.4% and 4.1% of the survivors were dis
charged with sequelae. The vast majority of those who died or developed dis
abilities were children after bacterial meningitis. The mortality rate of bacterial meningitis was 13% and the disability rate 10.3%. With the concentration of patients to the Tartu University Children's Hospital, the mortality as well as the disability rates decreased dramatically during these 10 years, which shows that the concentration of patients to the tertiary care hospitals is necessary.
REFERENCES
Abrikossova NYu, Kostjukova NN [Epidemiology of neonatal meningitis.]
Zh Microbiol Epidemiol Inf 1991; (2): 41-42 (Russ).
Adams WG, Deaver KA, Cochi SL, Plikaytis BD, Zell ER, Broome CV, Wenger JD Decline of childhood Haemophilus influenzae type b (Hib) disease in the Hib vaccine era. JAMA 1993; 269: 221-226.
Agrawal M, Bhandari NR. CSF glutamic oxalacetic transaminase levels in CNS infections. Indian Pediatr 1989; 26: 1245-1248.
Annegers JF, Hauser WA, Beghi E, Nicolosi A, Kurland T The risk of un
provoked seizures after encephalitis and meningitis. Neurology 1988; 38:
1407-1410.
Apak S, Kuremai N, Ozmen M Neurologische Dauerschäden bei bakterieller Neugeboren meningitis. Kinderarzt 1983; 10: 1234-1236.
Awapara M, Bhandari NR CSF glutamic oxalacetic transaminase levels in CNS infections. Indian Pediatrics 1989; 26: 1245-1248.
Bach F, Bach FW, Pedersen AG, Larsen PM, Dombernowsky P. Creatine kinase BB in the cerebrospinal fluid as a marker of CNS metastases and lep-tomeningealcarcinomatosis in patients with breast cancer. Eur J Cancer Clin Oncol 1989; 25: 1703-1709.
Baraff LJ, Lee SI, Schriger DL Outcomes of bacterial meningitis in children:
a meta-analysis. Pediatr Infect Dis J 1993; 12: 389-394.
Beaty HN, Oppenheimer S. Cerebrospinal fluid lactic dehydrogenase and its isoenzymes in infections of the central nervous system. N Engl J Med 1968;
279: 1197-1202.
Beghi E, Nicolosi A; Kurland LT; Mulder DW; Hauser WA; Shuster L Encephalitis and aseptic meningitis, Olmsted County, Minnesota, 1950-1981; I Epidemiology. Ann Neurol. 1984; 16: 283-294.
Bell AH, Brow D, Halliday HL, McLure G, McReid M Meningitis in the newborn — a 14 year review. Arch Dis Child 1989; 64: 873-874.
Belsey MA. CSF GOT in acute bacterial meningitis. Am J Dis Child 1969; 117:
288-292.
Belton NR Creatine phosphokinase (CPK), blood and CSF levels in newborn infants and children. Arch Dis Child 1970; 45: 600.
Bennhagen R, Svenningsen NW, Bekassy AN Changing pattern of neonatal meningitis in Sweden: a comparative study 1976 vs 1983. Scand J Infect Dis 1987;19:587-593.
Bhatt S, Halpin C, Hsu W, Thediger BA, Levine RA, Tuomanen E, Nadol JB Hearing loss and pneumococcal meningitis: an animal model. Laryngoscope 1991; 101: 1285-1292.
Botsvadze ES, Brteschvadze TT [Clinical characteristics and management of pneumococcal meningitis.] Pediatria 1983; (1): 59-60 (Russ).
Boyde TR Serum levels of mitochondrial aspartate aminotransferase in my ocardial infarction and muscular dystrophy. Enzymol Biol Clin 1968; 9: 385.
1 9
73
Briem H Cerebrospinal fluid in patients with meningitis and encephalitis.
Academic dissertation, Stockholm 1982.
Brookhauser PE, Auslander MC Aided auditory thresholds in children with postmeningitic deafness. Laryngoscope 1989; 99: 800-808.
Bryan JP, Silva HR, Tavares A, Rocha H, Scheid WM Etiology and mortal ity of bacterial meningitis in Northeastern Brazil. Rev Inf Dis 1990; 12: 708-715.
Bödvarson A, Franzson L, Briem H Creatine kinase isoenzyme BB in the cerebrospinal fluid of patients with acute neurological diseases. J Internal Medicine 1990; 227: 5-9.
Carter PE, Parclay SM, Galloway WH, Cole GF Changes in bacterial meningitis. Arch Dis Child 1990; 65: 495-498.
Chartrand SA, Cho CT Persistent pleocytosis in bacterial meningitis. J Pediatrics 1976; 88: 424-426.
Chatterley S, Sun T, Lien Y Diagnostic value of lactate dehydrogenase isoen
zymes in cerebrospinal fluid. J Clinical Laboratory Analyses 1991; 5: 168-174.
Cho HW, Prilipko L, Meitzer HY et al Isoenzymes of creatine phosphokinase in white blood cells. Experientia 1977; 33: 166-167.
Cibiras PP, Broslavski EI, Kelminskene SA [Etiology of bacterial meningitis in Lithuania.] Abstracts of the meeting Meningococcal infection and bacterial meningitis. Arhangelsk 1986: 45-47. (Russ)
Cizman M, Mozetic M, Radescek-Rakar R, Pleterski-Rigler D, Susec-Mitchieli M Aseptic meningitis after vaccination against measles and mumps.
Ped Infect Dis J 1989; 8: 302-308.
David DHO, Hirsch MS Acute viral encephalitis. Med Clin North America 1985;69:415-429.
Delanghe JR, Winter HAD, Buyzere LD, Camaert JJ, Martens FE, Praeter ML Mass concentration measurments of creatine kinase BB isoenzyme as an index brain tissue damage. Clin Chimica Acta 1990; 193: 125-136.
Devjatkina NP, Djomina AA, Piluschkov AV [Epidemiology of meningococ
cal infection in USSR.] Zh Microbiol Epidemiol Immonobiol 1989; (12): 36-40.(Russ).
Djomina AA, Iljina TA [Etiology of the bacterial meningitis today.] Zh Microbiol Epidemiol Immunobiol 1985; (3): 3-6. (Russ)
Djomina AA, Pokrovski VI, Iljina TV, Larina LI, Devjatkina NP, Markova AI [Etiology of the bacterial meningitis.] Zh Micrrobiol Epidemiol Immunobiol 1984; (12): 32-36.(Russ).
Dodge PR Sequelae of bacterial meningitis. Pediatr Infect Dis J 1986; 5: 618—
620.
Dodge PR, Davis H, Feigin RD Prospective evaluation of hearing impairment as a sequelae of acute bacterial meningitis. N Engl J Med 1984; 311: 869-874.
Etter CG; Wedgwood J; Schaad UB. Aseptische Meningitiden in Pädiatrie.
Schweiz Med Wochenschrift 1991; 121: 1120-1126.
Feigin RD Bacterial meningitis beyond the neonatal period, in R.D.Feigin and J.D.Cherry (eds.): Textbook of Pediatric Infectious Diseases, ed by 3rd ed.
Saunders-Company 1992: 401-428.
Feigin RD, McCracken GH Jr, Klein JO Diagnosis and management of meningitis. Pediatr Infect Dis J 1992; 11: 785-814.
Feldman HM, Michaels RH Academic achievement in children ten to 12 years after Haemophilus influenzae meningitis. Pediatrics 1988; 81: 339-344.
Feldman WE Cerebrospinal fluid lactic acid dehydrogenase activity. Am J Dis Child 1975; 129: 77-80.
Feld mann WE, Ginsburg CM, McCracken GH, Allen D, Ahmann P, Graham J, Graham L Relation of concentrations of Haemophilus influenzae type b in cerebrospinal fluid to late sequelae of patients with meningitis. J Pediatrics 1982; 100: 209-212.
Florez G, Cabeza A, Gonzales JM, Garcia J, Ucar S Serum and cere brospinal fluid enzymatic modifications in head injury. Fifth European Congress of Neurological Surgery. Oxford, 1975, Abstract 92.
Fortnum HM Hearing impairment after bacterial meningitis: a review. Arch Dis Child 1992; 67: 1128-1133.
Francis BM, Gilbert GL Survey of neonatal meningitis in Australia: 1987— 1989. Med J Austr 1992; 156: 240-243.
Friebel D, Henker J, Fritsche F Hörstörungen bei Kindern mit Meningitis purulenta durch Ampicillin. Kinderärzl. Praxis 1984; 10: 465^-69.
Fröber R, Müller G Epidemiologische und katamnestische Untersuchungen bei Meningitis purulenta im Kindesalter. Kinderärzl Praxis 1977; (5): 203-211.
Fuijnaga T, Mõtegi Y, Tamura H, Kuroume T A prefecture-wide survey of mumps meningitis associated with measles, mumps and rubella vaccine.
Pediatr Infect Dis J. 1991; 10: 204-209.
Gamstrop I, Klockhoff I Bilateral, severe, sensoneural hearing loss after Haemophilus influenzae meningitis in childhood. Neuropediatrics 1974; 5:
121-124.
Gary N, Powers N, Todd JK Clinical identification and comparative prognosis of high-risk patients with Haemophilus influenzae meningitis. Am J Dis Child 1989; 143:307-311.
Gilbert GL, Dickson KE, Waters MY, Kennet ML, Land SA, Sneddor M Outbreak of enterovirus 71 infection in Victoria, Australia, with a high inci dence of neurologic involvement. Ped Infect Dis J 1988; 7: 484—488.
Giustina D, Forese S, Pace P 20 anni (1960-1980) di meningite bacterica nell infanzia. Ped Med Chir 1985; 7: 195-202. (Italian)
Go KG, Patberg WR, Teelken AW, Gazendam J The Starling hypothesis of capillary fluid exchange in relation to brain edema. In Workshop on dynamic
Go KG, Patberg WR, Teelken AW, Gazendam J The Starling hypothesis of capillary fluid exchange in relation to brain edema. In Workshop on dynamic